Among India’s 11 crore registered equity investors as of early 2026, fewer than one in fifty can write down Benjamin Graham’s square-root intrinsic-value formula from memory. That single statistic explains why so much of the ₹1.81 lakh crore of cumulative F&O losses logged in the SEBI study fell on individuals chasing prices instead of valuing businesses. Graham published this formula in The Intelligent Investor in 1949 — seventy-seven years ago — and it still does the same work today: it converts two audited numbers from a company’s annual report into a back-of-the-envelope yardstick for what a defensive long-term investor might reasonably pay for a share. The formula is √(22.5 × EPS × BVPS). It is short, brutally honest about its assumptions, and almost completely ignored by today’s screen-trading retail investor.
What Is the Graham Number?
The Graham Number is the geometric mean of two earnings-quality anchors — a price-to-earnings cap and a price-to-book cap — multiplied by their respective per-share inputs. Graham observed that defensive investors had repeatedly been hurt by paying too much on either measure, and that combining the two into a single intrinsic-value ceiling forced discipline on both fronts at once. The full formula is:
Graham Number = √(22.5 × EPS × BVPS)
Where EPS is the diluted trailing twelve-month earnings per share (from the audited Statement of Profit and Loss) and BVPS is the book value per share (Net Worth ÷ Equity Shares Outstanding, from the audited Balance Sheet). The constant 22.5 is not arbitrary — it is the product of two ceilings Graham articulated explicitly in Chapter 14 of The Intelligent Investor: a maximum trailing P/E of 15× and a maximum P/B of 1.5×. Multiply 15 × 1.5 = 22.5. The square root of (22.5 × EPS × BVPS) is mathematically equivalent to saying: a defensive investor should not pay a price that would push both the P/E above 15 and the P/B above 1.5 at the same time.
Where the Formula Comes From — Graham’s Two Ceilings
Graham’s 15× P/E ceiling was rooted in his observation that the long-run earnings yield of the U.S. stock market between 1871 and 1947 averaged roughly 1/15 = 6.7%, comparable to the prevailing long-bond yield plus a modest equity-risk premium. In the Indian context of 2026, the 10-year G-Sec yield hovers in the 6.7–7.2% band, which is uncannily close to Graham’s reciprocal — meaning the 15× P/E ceiling translates almost directly to Indian markets without recalibration. Graham’s 1.5× P/B ceiling, similarly, was meant to ensure that the investor was not paying more than 50% above the net asset value disclosed in the audited balance sheet — a hard floor against goodwill-inflated or intangible-heavy stories that might evaporate in a downturn.
Graham allowed one disciplined relaxation: if a company’s P/E was below 15, he was willing to let P/B drift above 1.5 — as long as the product of P/E × P/B stayed below 22.5. That is the elegance of the square-root formulation. It captures the trade-off in a single value rather than enforcing two rigid ceilings independently.
How to Read the Graham Number for an Indian Stock — Step by Step
Pick any Indian listed company. Open its most recent annual report. Walk through these five steps:
Step 1 — Locate diluted EPS in the audited Statement of Profit and Loss. For Indian companies, this is mandated to be disclosed under Ind AS 33. Use trailing twelve months (last four quarters combined), not just the latest year, to smooth out one-off items.
Step 2 — Compute BVPS as Total Shareholders’ Equity (Equity Share Capital + Other Equity) divided by the number of equity shares outstanding at year-end. Subtract any intangible assets and goodwill if you want Graham’s stricter tangible-book version.
Step 3 — Plug into √(22.5 × EPS × BVPS). If EPS is negative, the formula fails — Graham would not have applied any valuation to a loss-making company. He required two consecutive years of positive earnings before computing his number.
Step 4 — Compare the result to the current market price. If the current market price is above the Graham Number, Graham himself would have said the stock is more expensive than what a defensive investor should pay. If it is below, it might qualify for further homework. The Graham Number is not a buy signal — it is a maximum-price screen.
Step 5 — Apply Graham’s seven defensive screens as preconditions. The Graham Number is meaningless on a company that fails any of these tests: adequate size; sufficiently strong financial condition (current ratio > 2, long-term debt < net working capital); earnings stability (positive earnings for each of last 10 years); dividend record (uninterrupted for 20 years); 10-year earnings growth of at least one-third; moderate P/E (under 15 trailing); moderate price-to-asset (under 1.5).
Two Contrasting Examples — Discipline vs Red Flag
The disciplined illustration: Consider a hypothetical Indian manufacturer reporting trailing diluted EPS of ₹45 and BVPS of ₹500. The Graham Number works out to √(22.5 × 45 × 500) = √(506,250) ≈ ₹711. A defensive investor reviewing the chart would treat ₹711 as the price beyond which Graham’s twin discipline starts to break. If the market price sits at ₹520, the formula leaves a margin of safety of roughly 27%; if it sits at ₹900, the formula warns the buyer that they would simultaneously be paying about 20× earnings and 1.8× book — outside both of Graham’s ceilings at once.
The red flag illustration: Consider a generic listed firm — there have been many examples across Indian small-cap history — reporting trailing EPS of ₹4 (slim margins) and BVPS of ₹35 (thin equity base, no retained-earnings build-up). The Graham Number would be √(22.5 × 4 × 35) = √(3,150) ≈ ₹56. If such a stock trades at ₹240 in the market, the implied multiples are roughly 60× earnings and 6.9× book — multiples that compress to fair value only if earnings grow at an exceptional clip uninterrupted for many years. Graham’s formula is not telling you the stock cannot work; it is telling you that you are paying the equivalent of three or four years of perfect execution in advance, with no margin of safety left if reality wobbles even slightly.
Titan Biotech FY25: What the Numbers Reveal
The Graham Number itself depends on per-share inputs that vary day-to-day with market cap and share count, so the responsible educational use of Titan Biotech here is not to declare a price target. Instead, this section illustrates the kind of audited quality discipline Graham would have insisted on as a precondition for even applying his formula. Below are the FY25 audited markers (twelve months ended 31 March 2025) drawn from Titan Biotech Limited’s Annual Report:
| FY25 Audited Marker | Reading | Graham Pre-Condition Satisfied? |
|---|---|---|
| Total Revenue FY25 | ~₹214 Cr (Q1 ₹46.5 Cr → Q2 ₹54 Cr → Q3 ₹56 Cr → Q4 ~₹58 Cr) | Sequentially growing top-line — Graham’s earnings-stability anchor |
| Total Borrowings FY25 | ₹3 Cr (essentially debt-free, D/E < 0.05x) | “Sufficiently strong financial condition” — Graham’s solvency filter |
| Contingent Liabilities FY25 | ₹7.78 Cr (low relative to net worth) | Conservative off-balance-sheet posture — supports BVPS reliability |
| CFO/Operating Profit FY25 | 103% (cash generation slightly exceeds operating profit) | Earnings quality — the EPS input to the Graham Number is fully cash-backed |
| Revenue 10-yr CAGR | ~15% | Comfortably above Graham’s one-third decade-growth minimum |
| PAT 10-yr CAGR | ~29% | Earnings-stability filter passed with substantial buffer |
| Gross Block / CWIP FY25 | ~₹57 Cr / ~₹11 Cr CWIP — Depreciation/Gross Block ~7% | Asset-base reinvestment supports forward earnings — Graham’s productive-asset test |
| Board Meetings FY25 | 14 (vs SEBI minimum of 4) | Governance discipline — Graham’s “honest management” stipulation |
| Export Revenue Share FY25 | ~34.5% across 60+ countries | Low customer concentration — earnings durability anchor |
Read together, these audited markers describe the kind of balance-sheet conservatism Graham looked for before he opened his calculator. The ₹3 crore total borrowings, the ₹7.78 crore contingent liabilities, and the 103% CFO/Operating-Profit conversion mean that the equity reported on Titan Biotech’s audited balance sheet is genuinely the shareholders’ equity — not a paper figure propped up by hidden obligations or non-cash accruals. The 14 board meetings against SEBI’s minimum of four, paired with an independent chairperson and an audit committee structured under Section 177 of the Companies Act 2013, reduce the governance-fraud risk that Graham repeatedly warned defensive investors about. The ~15% ten-year revenue CAGR and ~29% PAT CAGR satisfy Graham’s earnings-stability test by a comfortable margin.
Critically, none of this is a verdict on whether Titan Biotech is cheap or expensive on any given trading day. The market price multiplies a free-floating multiple on top of these audited inputs, and any honest application of the Graham Number requires the reader to look up today’s diluted EPS and BVPS for themselves. What the table above does is satisfy the seven defensive preconditions that Graham insisted must be true before the formula is even worth applying. This is the educational point: most retail investors plug numbers into the Graham Number formula on companies that fail Graham’s defensive screens, and then are surprised when the resulting “intrinsic value” misleads them. This is not a buy/sell recommendation on Titan Biotech Limited (BSE: 524717).
How Indian Retail Investors Should Actually Use the Graham Number
The Graham Number works best as a filter, not a target. Build a watchlist of 30–60 Indian small- and mid-cap companies that you understand qualitatively. Apply Graham’s seven defensive screens to each one — most will fail at least one screen, which is itself useful information. For the survivors, compute the Graham Number every quarter when results are released. Track the ratio of (current market price ÷ Graham Number) over time. When this ratio crosses below 1.0 for a company that has consistently passed the seven defensive screens, you have a candidate that deserves a deeper qualitative review — not an automatic buy, but a starting point for further work.
The 11 crore registered equity investors in India of 2026, up from approximately 3 crore in early 2019 per NSE data, are overwhelmingly first-generation participants who arrived during the post-pandemic rally. A simple Graham Number watchlist, refreshed quarterly using nothing more than the audited financials of the companies you already follow, is one of the cheapest disciplines you can adopt. It will not catch every multibagger — Graham himself missed many of the great twentieth-century growth stories because his formula penalised high-P/B technology businesses. But it will keep you out of the kind of structurally over-priced names that contributed materially to the ₹1.81 lakh crore of cumulative individual F&O losses documented in SEBI’s 2024 study.
Common Traps and Misinterpretations
Trap 1 — Using one-quarter EPS instead of trailing twelve months. Indian quarterly reports often contain seasonality or one-off items. Always annualise using last four quarters combined, and read the notes to flag any extraordinary item.
Trap 2 — Ignoring intangibles when book value is goodwill-heavy. If a company’s BVPS is inflated by goodwill from acquisitions, the Graham Number overstates intrinsic value. Compute tangible BVPS by subtracting goodwill and intangible assets from net worth before plugging in.
Trap 3 — Applying the formula to financial companies and software firms. Banks, NBFCs, and asset-light software companies do not fit Graham’s defensive framework. Graham himself excluded financials. For these, use sector-appropriate valuation tools instead.
Trap 4 — Treating the Graham Number as a price target. It is a defensive ceiling, not a fair-value estimate. A stock trading at half its Graham Number is not automatically attractive — it may have failed one of the seven defensive screens that you did not check.
Trap 5 — Forgetting Graham’s own disclaimer. Graham himself wrote in the 1973 edition of The Intelligent Investor that the formula was “presented as a useful approach, not a complete one.” Use it as one input among many, alongside cash-flow quality, governance, segment-mix, and qualitative judgement.
Key Takeaways
- The Graham Number = √(22.5 × EPS × BVPS) is a 1949 defensive-price ceiling, not a fair-value target. The 22.5 constant is the product of Graham’s 15× P/E and 1.5× P/B ceilings, designed to discipline both multiples simultaneously.
- The formula is only meaningful for companies that pass Graham’s seven defensive screens first — adequate size, strong solvency, earnings stability, dividend record, decade-long growth, moderate P/E, and moderate P/B. Plugging numbers into companies that fail any of these screens produces a misleading answer.
- Titan Biotech Limited (BSE: 524717) FY25 audited numbers illustrate the conservative-quality preconditions Graham looked for: ₹3 Cr total borrowings, ₹7.78 Cr contingent liabilities, 103% CFO/Operating Profit, ~15% revenue 10-year CAGR, ~29% PAT 10-year CAGR, 14 board meetings, ~34.5% export share across 60+ countries. This is an educational illustration of the kind of balance-sheet discipline that satisfies Graham’s pre-valuation filters — not a buy/sell recommendation on the stock.
- For India’s 11 crore equity investors in 2026, a quarterly Graham Number watchlist is one of the cheapest and most underused discipline tools available. It will not catch every multibagger, but it will keep you out of structurally over-priced names that contributed to the ₹1.81 lakh crore of cumulative individual F&O losses documented in SEBI’s 2024 study.
SEBI Disclaimer
Disclaimer: This article is for educational and informational purposes only. It is not investment advice, and not a buy, sell, or hold recommendation on any stock mentioned, including Titan Biotech Limited. Equity markets carry risk; please do your own research or consult a qualified professional before making investment decisions.
